The rapid expansion of Internet of Things (IoT) applications necessitates the effective deployment of base stations (BSs) to enable consistent connectivity across large geographic areas under interference-limited conditions. Existing techniques typically use distance-based or binary coverage models; however, these abstractions fail to account for the influence of co-channel interference on the quality of communication in dense deployments. In this paper, we investigate the Signal-to-Interference-plus-Noise Ratio (SINR)-aware Base Station Deployment (BSD) problem in wide-area IoT sensor networks. The objective is to determine a minimum-cost subset of BSs from a predefined set of candidate BSs such that every IoT sensor is covered by at least one BS and a target SINR threshold is satisfied. The problem is formulated as a combinatorial optimization problem, which is NP-hard. Theoretical analysis establishes that the proposed coverage function is monotone and submodular, enabling the SINR-aware greedy algorithm to achieve a (1-1/e)-approximation to the optimal solution while maintaining a polynomial-time computational complexity. Numerical evaluations on a real water distribution network dataset demonstrate that the proposed SINR-aware greedy algorithm achieves near-optimal base station deployment while significantly reducing computational effort. Compared with the Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) algorithms, the proposed approach attains complete sensor coverage with deployment costs within 12.3% of the best-performing metaheuristic solution while requiring up to 190 times lower execution time.

翻译：物联网应用的快速扩展要求在干扰受限条件下，跨广阔地理区域有效部署基站（BS）以支持持续连通性。现有技术通常采用基于距离或二元覆盖模型，但这些抽象方法未能考虑同信道干扰对密集部署中通信质量的影响。本文研究了广域物联网传感器网络中面向信干噪比（SINR）感知的基站部署（BSD）问题。其目标是从预定义的候选基站集合中，确定一个成本最低的子集，使得每个物联网传感器至少被一个基站覆盖，且满足目标SINR阈值。该问题被形式化为一个组合优化问题，属于NP困难问题。理论分析表明，所提出的覆盖函数具有单调性和子模性，使得SINR感知贪婪算法能在保持多项式时间计算复杂度的同时，实现(1-1/e)最优解的近似。基于真实供水管网数据集的数值评估显示，所提出的SINR感知贪婪算法在显著降低计算开销的同时，实现了接近最优的基站部署。与遗传算法（GA）和粒子群优化（PSO）算法相比，该方法在部署成本不超过最优元启发式方案12.3%的情况下，实现了完全传感器覆盖，且执行时间最多降低190倍。